A company is its employees. How those employees interact with each other and what shared environment they create are among the most important aspects of any company (and among the first things a smart investor or potential hire will look at). Another word for this is culture, and it matters.
It makes perfect sense, then, that hiring managers attempt to judge the cultural impact of each applicant. A single toxic employee who is a bad fit for the existing culture can tank the morale and productivity of an entire team. Screening for this is entirely logical.
In practice, however, culture fit screens have a dark side. They are typically conducted free-form, where an interviewer talks to a candidate for 45 minutes and makes a gut call. Few companies clearly enumerate what traits they look for, let alone how to evaluate them consistently. This opens the door to bias. The human brain is lazy, and in the absence of clear evaluation criteria, we substitute pattern matching. And because the culture fit screen holds a veto over other parts of the interview, this bias goes unchecked. Poorly executed, a culture fit screen can be the most capricious section of an interview.
What do we do, then? Culture fit is important, but attempts to screen for it can introduce noise and bias. The answer, I think, is to further dissect the concept. By breaking down culture fit, it becomes easier to design interview questions that get at the parts that matter and avoid the dark side. My goal for this blog post is to do this. Over the last year, I've spoken to hiring managers at over 300 tech companies . I asked each how they define their culture, and how they evaluate fit. Their answers were the starting point for this post.
What is culture fit?
Let's start by defining the term. Interestingly, companies themselves often find this hard to do! When I asked companies to define their culture, most struggled at first to give clear answers. Still, over all the conversations, the following 3 categories emerged:
- Communication and soft skills
- Specific personality traits
- Gut feel / friend test
I think this is a useful way to breakdown the concept, and I'm going to dig into each.
By 'soft skills', I mean a candidate's ability to communicate clearly, take ownership, and remain positive and friendly. These are things that nearly every company values! Critically, soft skills are NOT really about 'fit' - that is, they're not specific to a particular company. Rather, they are general, widely valued non-technical skills.
The fact that all companies value soft skills would seem to take them outside the scope of what is usually meant by company culture. After all, they don't distinguish any one company. They are just skills we all agree make a better employee (like, say, intelligence and work ethic).
However, communication ability, ownership and positivity were the #1, #2 and #3 most mentioned attribute when hiring managers talked to me about their culture. In fact, well over half of companies look at only these three things when judging culture fit . A large portion of what culture fit means in practice is far less diaphanous than the term implies. It's largely a matter of three common soft skills.
The takeaway for candidates, I think, is that they can focus their interview preparation. Beyond technical skills, what matters is communication, ownership and positivity. If these are things you struggle with, you need to practice them. You can read more about our advice for preparing for technical interviews here.
The takeaway for hiring managers is that they should consider running more structured culture fit screens. Soft skills are likely already a large part of what your screen looks at. And there's a wealth of evidence that targeted assessments outperform unstructured interviews, and leave less room for bias. You should consider developing interview sections directly targeting communication, ownership and positivity, or at least using scorecards that ask interviewers to rate candidates on these three areas. Simply asking interviewers to rate candidates in specific areas that matter to your business (rather than asking them to judge the candidate as a whole) does a lot to reduce bias.
Specific personality traits
Soft skills are not the only thing that companies look for during culture fit screens. Some companies also look for specific personality traits that more traditionally fit under the label of culture. For example, Bridgewater Associates believes in radical honesty, and encourages employees to critique each other's flaws. They screen out candidates who are uncomfortable with this. Stripe has a company-wide focus on friendliness, and screens out candidates who most other companies would hire. And Uber, famously, gives (gave?) preference to don't-take-no-for-an-answer candidates. A key point is that these companies are screening out otherwise qualified candidates who have generally good soft skills, based on personality traits specific to the company culture they're targeting.
There are also plenty of companies, however, that don't screen for traits beyond soft skills. Facebook? Microsoft? Dropbox? I'm sure these companies have distinctive cultures. But they don't impose strong cultural filters when evaluating candidates. They hire reasonably friendly people who can do the work. Interestingly, this actually appears to be the norm. Among the 300 companies I spoke to, only 20% told me they engaged in screening for specific traits beyond soft skills.
Looking at our data, I see no correlation between whether a company screens on personality traits and whether they are successful. From the data alone, we can't conclude much. However, it is my opinion that most companies should not engage in this type of screening.
A company is its employees. Finding enough strong people to join your company is itself a huge challenge. And if you reject otherwise strong candidates because of a cultural thesis, you are making that challenge harder. If you're even a little bit wrong about that thesis, you pay a heavy cost.
Obviously, screening for specific personality traits has not kept Bridgewater or Stripe from succeeding. Uber, however, might be a different story. I am going to argue that personality trait screening may have harmed Uber. They experienced a lot of turmoil last year, with issues stemming from an aggressive culture. For years before this, they selected for aggressive employees, who wanted to be owners, not renters. My claim is that Uber's problem was not insufficient attention to culture. Rather, it was too much attention to culture. Had they hired for soft skills and ability to do the job, and not explicitly selected the most aggressive people, they might have been in a better place.
If you're a hiring manager and you do want to screen on personality traits, I would advise you to think about how you communicate what you are screening for. Companies that want to do this tend to produce a list of qualities they want in an ideal employee. The problem with this is that, inevitably, the list grows to include mutually exclusive traits, and it becomes unclear who exactly is being screened out. If you want to screen on a personality trait, you will be screening people out. I recommend you answer the question,
who are the otherwise-strong candidates you don't want to hire.
Gut feel / friend test
The final category of culture fit I want to discuss involves making a gut call on whether a candidate feels like someone who would fit in at your company. Sometimes this is phrased as judging whether a candidate is someone you could be friends with. This is the most dangerous of the types of culture fit screening. Once you account for soft skills, whether a candidate feels like someone you'd like to be friends with says more about you and your past friends than it does about that person.
To see this more clearly, think back to the first time you met your spouse or partner. What was your impression of them? How similar is this to how you now view them? Most of the time, our initial impressions of people are highly inaccurate. Research into hiring bears this out. Unstructured initial assessment has little correlation with performance. Pair this with the fact that humans show a friendship preference for people of the same age, gender and race, and the friend test comes into focus as a recipe for bias .
The good news is that, among the companies I spoke to, it was also the least common. Less than 10% of companies do this sort of screening (at least explicitly).
Company culture matters. However, unstructured culture fit screening can be a recipe for bias. The solution is to break the screens down into three types of assessment. Soft skills are important, and companies should set a high bar here. They should use structured interviews to measure these skills. Screening on specific personality traits may make sense for some companies. However, doing it comes with a high cost (rejecting otherwise strong candidates). Screening candidates on how they make you feel and whether they seem like someone you could be friends with is a mistake. It's poorly correlated with job performance, and invites bias.
I'm not the first person to write critically about culture fit. However, the pushback to this criticisms is always that culture is really important, and we can't just stop screening for it. What I think I've uncovered here is that these two arguments are not actually in conflict. Soft skills are really important. We can't stop screening for them. But we can screen for them more effectively, and without introducing the bias of unstructured culture fit interviews. I think it might be time to retire the word 'culture fit', and just talk about soft skill screening.
I've been thinking about this a lot. If you have any thoughts, send me an email at email@example.com. And if you're interested in getting a programming job without navigating the maze that is culture fit, you should try the Triplebyte process.
 Part of bringing a company onto Triplebyte used to be an in-depth discussion of the company's hiring needs, including a discussion of culture fit. I did 300 of these calls. We've recently replaced this with an improved product, where companies can indicate their hiring preferences without talking to me.↩
 The notes that companies send Triplebyte after they interview a candidate also back this up. When companies tell us they failed someone for culture fit reasons, usually the detailed notes mention one of these things.↩
 A case is sometimes made for the friend test, when a startup is hiring its first few employees. The argument is that a startup where the early team is friends has an advantage. There might be some truth to this. But it's still a really dangerous way to screen people. I think even for early stage-startups this should be avoided. ↩